Re: [PATCH] sctp: check transport existence before processing a send primitive

By the way, I'm thinking of setting err to -EAGAIN here so that userspace can retry the send and sctp_sendmsg() will try again to find a suitable association or create a new one if necessary, but if someone more knowledgeable about the SCTP implementation has a better suggestion about what kind of error to return in this scenario, I'd appreciate it.

Regards, Ricardo